1. Technical Field
This invention relates in general to managing power consumption in a high availability system and more particularly to managing power savings in a high availability system at a redundant component level of granularity.
2. Description of the Related Art
In some computing environments, it is critical that the computing environment continue to handle workloads even if one or more resources handling the workloads within the computing environment, fail. For a computing environment to continue to handle workloads, even if one or more resources handling the workloads within the computing environment fail, the computing environment may implement a high availability (HA) system including redundant computers in groups or clusters and a HA controller that provides for automated continued service to workloads when system components within the computer environment fail.
In an HA system, redundant computers include at least one primary node with one set of components and at least one standby or backup node with a backup copy of the set of components in the primary node. To provide high availability to workloads, when system components fail or other conditions in the cluster change, the HA controller detects when the conditions in the cluster change and moves the resource group for a workload to a standby node to continue working. Since failures are unpredictable, in an HA system, the standby node remains powered on, waiting in a failover mode for a failure. Thus, in an HA system, the overall power consumption in a configuration that runs at least one standby node for each primary node, is higher than in a system that is not configured for high availability.